gavreid
Pretty Words...
Correct me if I’m wrong, but that looks as though a line with roughly half the gradient would still fit those new data points, albeit the data are a bit noisier than before (possibly in consequence of the smaller number of data points?). That would, I think, imply roughly a doubling of the time period to halve, but still an exponential curve? In any event, the trend is still downwards, and meaningfully so and, as you say, it’s well below where it would have been, pre-vaccination.
You can do that but it's just fitting for the sake of fitting - it isn't the same exponential as you say. The red line is the same as we've had since January, I haven't changed it for precisely this moment in time, now we can easily compare. There are really two competing effects the flattening and now the probable increase in cases and the race to vaccinate.